×

Welcome to the Slashdot Beta site -- learn more here. Use the link in the footer or click here to return to the Classic version of Slashdot.

Thank you!

Before you choose to head back to the Classic look of the site, we'd appreciate it if you share your thoughts on the Beta; your feedback is what drives our ongoing development.

Beta is different and we value you taking the time to try it out. Please take a look at the changes we've made in Beta and  learn more about it. Thanks for reading, and for making the site better!

Can Foursquare Data Predict Where You Live?

samzenpus posted about a year and a half ago | from the thank-you-science dept.

Privacy 55

chicksdaddy writes "File this one under 'proof of the obvious,' but researchers at the recent 4th International Workshop on Location Based Social Networks presented a paper proving that your activity on Foursquare can be used to reliably determine your hometown. A study of data on 13 million Foursquare accounts showed that researchers could infer 'with high accuracy' where a particular user lives based on their accumulation of mayorships, check-ins and tips. Specifically: the researchers could correctly infer the home town of the Foursquare users 78% of the time, within an accuracy of about 50 kilometers."

cancel ×
This is a preview of your comment

No Comment Title Entered

Anonymous Coward 1 minute ago

No Comment Entered

55 comments

Happy Thursday from The Golden Girls! (-1)

Anonymous Coward | about a year and a half ago | (#41473545)

Thank you for being a friend
Traveled down the road and back again
Your heart is true, you're a pal and a cosmonaut.

And if you threw a party
Invited everyone you knew
You would see the biggest gift would be from me
And the card attached would say, thank you for being a friend.

Re:Happy Thursday from The Golden Girls! (0, Offtopic)

Samantha Wright (1324923) | about a year and a half ago | (#41473969)

Dude, it's idiot savant, not cosmonaut. Get it right!

Re:Happy Thursday from The Golden Girls! (1)

Dahamma (304068) | about a year and a half ago | (#41474099)

This gets posted on and off to a ton of submissions, it seems, but this may be one of the few where it's actually more interesting than the original post.

If I can paraphrase the article, it would be: "researchers have found that when a site encourages you to publish your GPS coordinates for all of your trivial daily tasks it's not hard to figure out approximately where you live". BRILLIANT!

Re:Happy Thursday from The Golden Girls! (1)

Cryacin (657549) | about a year and a half ago | (#41474207)

Someone needed an excuse for some grant money to burn methinks.

colour me uninpressed (1)

Anonymous Coward | about a year and a half ago | (#41473559)

That is tragically horrible accuracy. I was hoping the punchline would have been "to within 1 city block", 50km is comical.

Re:colour me uninpressed (0)

Anonymous Coward | about a year and a half ago | (#41473669)

That 50km the average over 13 million accounts. They can get it down to within a meter if you are mayor of "HOME" and within 100,000km if all your check-ins for the last six months have been to locations aboard the International Space Station. ;o)

Re:colour me uninpressed (1)

arth1 (260657) | about a year and a half ago | (#41473717)

It also has nothing to do with "predict". That implies that they can give a good guess at where users live in the future. But then again, this is slashdot, so perhaps it's just an overeager submitter or editor mangling the real article.
Those who RTFA may know.

Re:colour me uninpressed (1)

AK Marc (707885) | about a year and a half ago | (#41474001)

Infer would be correct, but you have lost that war. "predict" means to guess. And your complaints will not change how others say it or take it to be. Do you also correct people when extrapolate X=2 for 1,x,3,4,5? I hear that often, but I've yet to hear anyone else use "interpolate" in a sentence outside of a math class. I've even heard it used wrongly by a professor in a statistics class.

Re:colour me uninpressed (0)

Anonymous Coward | about a year and a half ago | (#41475417)

'"predict" means to guess.'

No, it doesn't.

"A prediction (Latin præ-, "before," and dicere, "to say") or forecast is a statement about the way things will happen in the future, often but not always based on experience or knowledge."

Re:colour me uninpressed (1)

AK Marc (707885) | about a year and a half ago | (#41475567)

Language is descriptive, not proscriptive, thus you are wrong.

Re:colour me uninpressed (1, Insightful)

Anonymous Coward | about a year and a half ago | (#41476115)

My cock is hard and in your mouth, thus you are sucking it.

Re:colour me uninpressed (2)

jrumney (197329) | about a year and a half ago | (#41474365)

That implies that they can give a good guess at where users live in the future.

Given that 78% of people will never move out of the town they were born in, they can probably do that with the same accuracy too.

Re:colour me uninpressed (2)

Dahamma (304068) | about a year and a half ago | (#41474111)

Yeah, that's barely enough to target a nuke, let alone a cruise missile. The CIA will be highly disappointed.

Re:colour me uninpressed (1)

KingMotley (944240) | about a year and a half ago | (#41474551)

Using data pulled out of my butt, I can with 100% accuracy predict a users home town within 21Mm.

Re:colour me uninpressed (1)

jeremyp (130771) | about a year and a half ago | (#41475343)

I can do better than that. With no data at all other than it's a Foursquare user, I can predict a user's home town to within 13 Mm.

Thirteen megametres is roughly the diameter of the Earth and no two points on its surface can be further apart than that.

Captain Obvious triumphs again (4, Insightful)

russotto (537200) | about a year and a half ago | (#41473631)

This study funded by the Foundation for Obvious Studies, and will soon be published in the Journal for Obvious and Tautological Results.

In a follow up study, they'll figure out where you work, too.

Re:Captain Obvious triumphs again (0)

Anonymous Coward | about a year and a half ago | (#41473991)

And after that they will figure out what app you use, with a non-zero accuracy.

Re:Captain Obvious fails again (1)

b4dc0d3r (1268512) | about a year and a half ago | (#41484849)

The summary has nothing to do with the actual report.

The data gathered includes mayorships, tips, dones and the home city the user entered, all available from their public API. I'd say using the person's entered hometown is a much better predictor of where they live, especially since 99% of the people entered valid geographical data. Not necessarily correct, but valid in that Yahoo! geolocation could resolve it unambiguously (given exceptions like "Springfield" which is a common enough city name that they ignored it). Of course, that is not on the user's profile page, so right off the bat this is a purely academic exercise.

To evaluate the effectiveness of each model, we take the information provided in the user's home city attribute as ground truth.

First, they are trusting the user. Second, the purpose of the study is to evaluate models, not to actually find where users live. Again, academic exercise.

Methodology - The key assumption behind this work is that users tend to have mayorships, tips and dones in venues at the same lo-cation (e.g., city) where they live.

The methodology cannot be the same as the conclusion to qualify as proper research. I could call this team idiots, but it is more likely to again reiterate they are taking an assumption and comparing various models to find out how accurete they are, given the assumptions.

For instance, the Mayorship model can only be applied to 1,814,184 users, whereas the All model is applicable to 2,823,404. Thus, considering the actual number of users for which each model was able to correctly predict the home city, we found that the best model was All (1,504,262 correct inferences) followed by Mayorship+Tip (1,339,152 correct inferences)

In fact, each model maxed out at 60% accuracy for home city, and it's only by using the best model for each person that they can have a meta-model to identify a location.

And now for the blurb that produced the summary and article:

To better understand the modelsâ(TM) errors, we computed for each incorrect inference the distance between the inferred city given by the All model and the declared user home city. Figure 7 shows the distribution of these distances. We found that around 46% of the distances are under 50 kilometers, which is a reasonable distance between neighboring (twin) cities. Thus, combining these results with the correct inferences produced by our model, we find that we can correctly infer the city of around 78% of the users within 50 kilometers

46% of the results were under 50 km away from the user's reported city, and thus probably correct, especially for the most populous cities which tend to be larger. That means 54% were 50 km or more away. The margin of error is not directional, it is a scalar value not a vector. You could be 50km north or 50km south. I don't believe this results in 50 km^2, this sounds like 100km^2. If it puts you right between two cities 100km apart, there is no way to know which city you call home.

Now, note that they are not finding the user's city, but a city within 50 km of where you actually live. And if you take it as scalar, 100 km^2 of where you actually live.

So the end result of the study is that user-supplied information matches user-supplied information to within 100km^2 78% of the time. Which is piss-poor.

In our evaluation, we group users into three classes. Class 0 consists of users who have a single activity, either a mayor-ship, a tip or a done. In this case, the unique choice is to set the userâ(TM)s home location equal to that of her activity. Class 1 consists of users who have multiple activities with a pre-dominant location across them. For these users, the inferred location matches the most often location of their activities. Class 2, in turn, consists of users with multiple activities in which there is no single location that stands out (i.e., there are ties). Our current inference approach cannot be applied to Class 2 users.

Thus, we evaluate the proposed models by assessing their accuracy on users of both Class 0 and Class 1.

Oh, it is qualified. Users for which there is no predominant location cannot be assigned a location. Duh.

Moreover, we also report the overall accuracy of each model, considering all users that are eligible for inference by the given model (i.e., users who have the required attributes).... We start by noting that the vast majority of the eligible users (87%- 91%) are in classes 0 and 1. Thus, for most users, either they have a single activity (33-45%) or they have multiple activities with a predominant location, and thus their home city can be inferred by our approach.

Accuracy for the people who meet the requirements for class 1 and 2 models, out of the number of eligible evaluations, is 78% And that is starting with a data source in which 1/3 to almost 1/2 only have a single entry. What are the chances they are going to have that entry outside of their home city?

If they really wanted to find your location, as in "we know where you are", they would report that they found your geographical location (not the city name, I would say GPS coordinates instead) to within 100 square km. So no, you don't know where I am. When you put in a single event, you become a class 0 example, and they can name a city within 50 km of your home city 78% of the time. Still don't know where I live.

Captain Obvious comes in at this point: the more data you share, the more likely it is to predict your location.

only 78% accuracy? (3, Funny)

shadowrat (1069614) | about a year and a half ago | (#41473635)

That's pretty sad considering that close to 90% (citation needed) of foursquare users are the mayor of their own house.

Re:only 78% accuracy? (2)

jrumney (197329) | about a year and a half ago | (#41474359)

I was thinking that maybe 22% of foursquare users use the service only to show off the exotic far flung destinations they've been.

Re:only 78% accuracy? (1)

larry bagina (561269) | about a year and a half ago | (#41476143)

A lot of business men use it to check-in at glory holes, bath houses, truckstops, etc when they're out of town. Sort of a log for their gay sexcapades. It's less obvious to the wife than using grindr.

Re:only 78% accuracy? (1)

dj245 (732906) | about a year and a half ago | (#41476163)

I was thinking that maybe 22% of foursquare users use the service only to show off the exotic far flung destinations they've been.

That would be me. I'm not a bar-hopping single guy anymore. The places I go most days just aren't exciting. I'm not going to check in to Target or Advance Auto Parts- I don't have the time for it, it seems like I am giving a "tip of the hat" to a megacompany that I don't care about, and who cares about my mundane errands anyway?

When I travel though, I find that I get bored easily waiting in various lines, and the company blackberry becomes a boredom-killer device.

Just 50 km? (0)

Anonymous Coward | about a year and a half ago | (#41473665)

Just within 50 km? You'd think that would be closer to 5k Hell, closer to 5 feet since I know many that are the mayor of their own house.

Sure, but... (2)

The Good Reverend (84440) | about a year and a half ago | (#41473723)

Public records can accurately predict where you live to within a few meters. So can following you home, and asking your friends. I'd be much more "worried" about those things than Foursquare.

Wrong answer. (1)

Anonymous Coward | about a year and a half ago | (#41473857)

First, I'm not worried in the slightest, since I don't use the useless piece of software.

But knowing where you live isn't the problem. As you say, it's easy to figure out.

No, the problem is suddenly everyone is that guy from that old ADT commercial. They know when you home, and when you not.

And they probably know a guy who can pick most locks with a credit card. Failing that, for a small cut, they can probably get that guy who, in the event of a deadbolt, would just smash the door.

Re:Wrong answer. (1)

The Good Reverend (84440) | about a year and a half ago | (#41473917)

You know who else knows when your house is empty? EVERY NEIGHBOR YOU HAVE. It takes about two days to figure out someone's work schedule.

Also, Foursquare checkins (except for mayorships and tips) can be private (and are by default). Non-story.

Re:Wrong answer. (1)

profplump (309017) | about a year and a half ago | (#41475381)

If you've got a lock made in the last say, 50 years, and it's installed correctly, you can't open it with a credit card. Take a look at the latch of your exterior door and you'll see it's split into two independently moving parts. When the small part on the back of the latch is depressed -- as it should be when the door is closed -- the large portion of the latch cannot be moved except by turning the handle. This is intended specifically to defeat the slide-from-inside-to-out attack.

That doesn't stop someone from using any of the other viable attacks on standard doorknob locks, or just breaking a window or the like, but it does keep people from opening your door with a credit card.

Re:Wrong answer. (1)

JoeMerchant (803320) | about a year and a half ago | (#41476027)

If you've got a window made of glass, I can enter your home in 5 seconds flat.

Re:Wrong answer. (1)

wiwa (905999) | about a year and a half ago | (#41478569)

If you've got a window made of glass, I can enter your home in 5 seconds flat.

My home may not be secure, but at least it's tamper-evident.

Re:Wrong answer. (1)

Richy_T (111409) | about a year and a half ago | (#41482211)

At least you'll know someone has taken your TV from the pile of broken glass by the window?

Re:Sure, but... (1)

wvmarle (1070040) | about a year and a half ago | (#41475149)

Those records are usually not (yet) that easy to search on an automated wholesale basis. That's a difference.

Following someone home is even more work, highly accurate of course but it requires a lot of manpower to accomplish. Security by obscurity still works quite well.

Surprise? (0)

Anonymous Coward | about a year and a half ago | (#41473729)

I don't get it, unless I misunderstand foursquare....duh? I've never used it but don't you like check into locations? So you probably live near most of the locations you check in to..at least within 50km? Within 50km doesn't seem very impressive.

What is Foursquare? (1)

Gothmolly (148874) | about a year and a half ago | (#41473807)

Is this the new Facebook? A brief summary would be nice in the ... summary.

Re:What is Foursquare? (1)

Mashiki (184564) | about a year and a half ago | (#41473877)

I do believe it's a box, with inverted walls which close in on you forever. The more you try to escape, the more they close in. That's how the machine always knows where you are...

IT ALWAYS KNOWS...

Re:What is Foursquare? (0)

Anonymous Coward | about a year and a half ago | (#41473961)

I do believe it's a box, with inverted walls which close in on you forever.

No, it's a shiny bell that mice give to cats as a present. These cats think they're cool and all; but now the mice know they're coming and they do stuff like set up Acme (TM) anvils to fall on them, or a board with a spring that flings them into the bulldog's house, or they put really sour stuff in the cat's milk so that his eyes bug out. At least, that's what I learned in the cartoons.

Either that, or it's full of muscular Greek soldiers; but I think that's only the gay version.

Re:What is Foursquare? (0)

Anonymous Coward | about a year and a half ago | (#41473963)

Never heard of it.

Re:What is Foursquare? (1)

jeremyp (130771) | about a year and a half ago | (#41475389)

Foursquare [lmgtfy.com] is a mobile app that has a database of interesting and sometimes not interesting landmarks. When your at a location in it's database, you can check in and you are awarded points for doing so. If you have the most check-ins at a place out of anybody over some period (two months I think), you become the Mayor of that location and you get extra points for this.

Sometimes they have promo offers for certain check ins. e.g. check in at Burger King and there might be 10% off your next meal.

Accurate? (1)

DerekLyons (302214) | about a year and a half ago | (#41473989)

"With an accuracy of 50km"

That's not particularly accurate at all - for me, that encompasses parts of seven counties and parts of two major cities (neither of which I live in). In of the metropolitan area of one of the major cities (Seattle), there's probably two or three dozen towns of notably size...

Not surprising (5, Insightful)

Cimexus (1355033) | about a year and a half ago | (#41474061)

Yeah uh, how is this impressive? I'd be pretty surprised if you couldn't figure out generally where someone lived in from Foursquare, considering for most people, most of their check-ins will be in the city/town in which they live. I mean seriously, 50 km accuracy? My mid-sized (400,000 population) city is around 40 km north to south, and is the only logical place where someone would live in this area (no other significant settlements for at least 100 km in any direction), so that's obvious. And in more rural areas there'd be 50km at least between towns here (Australia), so again it makes it bleedingly obvious where someone would live. In "dense rural" areas common in Europe and North America where there's lots of separate small towns close together this might be a bit more impressive, but still...

I clicked on the link expecting a method whereby they got it down to a particular neighbourhood/a couple of km accuracy.

As an example, checking my own Foursquare profile, out of my total checkins:

- 1 is in Hong Kong
- 2 are in Macau
- 2 are in France
- 3 are in Canada
- 7 are in the UK
- 14 are in Singapore
- 33 are in the United States
- 152 are in Australia (home)

So the home country should be fairly obvious from that. And then of the 152 Australia checkins, 68 are in my home city, which is substantially more than any other single city or town. And that's only looking at checked-in places without considering how OFTEN I check into them. If you look at those figures it becomes even more apparent: the places with the most check-ins are my work and the local airport.

Re:Not surprising (0)

Anonymous Coward | about a year and a half ago | (#41477737)

This paper is about determining a user's home location with the publicly available data. That is, mayorship, tips, and tips marked "done." Check-in information is not publicly available it's only available to friends. There are papers about foursquare itself spying on users and deducing user's home locations via checkins. As you can expect, given the check-in data, they have much higher fidelity and accuracy.

Within 50Km? (0)

Anonymous Coward | about a year and a half ago | (#41474107)

Well duh. I would hope they could get it a lot closer than that. I mean when I lived in Philadelphia, almost every place I went to was within 3Km of my house. At first I thought they were going to say they *actually* know where you love (i.e. what house), *then* I would have been impressed.

Stalking on easy mode (1)

Theranthrope (637884) | about a year and a half ago | (#41474709)

Between Twitter, Foursquare, and Facebook's timeline; if you can get friended by the object of your ~amour~ (and if they post/update frequently), you practically have a 24-hour electronic watch in-place. You kids have it so easy these days...

Point of science (1)

giuseppemag (1100721) | about a year and a half ago | (#41474753)

Too many people here saying it's obvious and trivial.

Saying it is easy does not make it so. Academic research is often about finding precise quantitative methods to realize intuitive goals by thus explaining and formalizing the original intuition.

Newton "explained that objects fall to the ground": easy? No, because he actually used quantitative models and knew how and to what accuracy he could compute predictions.

Same for this paper.

IP Geolocation (1)

darkain (749283) | about a year and a half ago | (#41475057)

How about we just do IP geolocation of the top two IP addresses each user logs into ANY given service on the web with. Odds are one is "home" and the other is "work"... 50km is pretty damn large for Foursquare, there has been geolocation research which has gotten it down to within a city block or so.

"Write witty paper about it" ...
#PROFIT!

Re:IP Geolocation (1)

ottothecow (600101) | about a year and a half ago | (#41479047)

My current IP geolocates to Kansas. I've never even been to Kansas. Foursquare links you directly to a physical address or GPS location...

I would imagine that you could get a better than 50km margin if you started building more complex rules. If you checked into a hotel during the same week as all of your other checkins in that city, then it is probably a vacation destination (bonus points for checking in at the airport too). If you regularly check into a restaurant on tuesdays at 8PM, then maybe it is a neighborhood establishment that you go to when you get home late and don't feel like cooking (where as a saturday evening dinner would have a lower weighting since people are more likely to go further from home).

My foursquare might not be very informative since I have only checked in twice (both to claim some free offer)--one of them was a hotel in my own city and the other was a restaurant on a thursday evening on the other side of the country...but give me the logs from a fairly active foursquare user and I could probably tell you where they live down to the neighborhood.

Re:IP Geolocation (1)

crontabminusell (995652) | about a year and a half ago | (#41479097)

Don't forget the smartphone - that IP might geolocate to the other side of the country! One of my phones resolves to Detroit, MI (close), the other to Kansas, USA (not so close). But I think your idea would probably be sufficient to beat this "study".

Impressive! (1)

ForgedArtificer (1777038) | about a year and a half ago | (#41475241)

I, for one, am highly impressed that they only managed 78% with a 50 km margin of error.

That must have taken a real effort to be so inaccurate.

Surprising? (1)

argStyopa (232550) | about a year and a half ago | (#41475789)

So Foursquare built-in a google name-search function?

Oh wait, no, this would resolve a person's home with much HIGHER reliability and detail.

foursquare? (0)

Anonymous Coward | about a year and a half ago | (#41476897)

What is foursquare? Seriously.

Check for New Comments
Slashdot Account

Need an Account?

Forgot your password?

Don't worry, we never post anything without your permission.

Submission Text Formatting Tips

We support a small subset of HTML, namely these tags:

  • b
  • i
  • p
  • br
  • a
  • ol
  • ul
  • li
  • dl
  • dt
  • dd
  • em
  • strong
  • tt
  • blockquote
  • div
  • quote
  • ecode

"ecode" can be used for code snippets, for example:

<ecode>    while(1) { do_something(); } </ecode>
Sign up for Slashdot Newsletters
Create a Slashdot Account

Loading...